The movie "The Minority Report" is set in a future where crimes can be detected before they happen and people are arrested and charged for acts they intended to commit. Tom Cruise heads up the Pre Crime unit that is powered by psychic beings whose dreams of the future can detect murderous intent, however predicting crime before it happens may not require such a far fetched premise.
Some would argue that security cameras do not prevent crime they only assist with catching the individuals who conducted those criminal activities. While they do provide invaluable information about perpetrators the police can use to quickly apprehend suspects, new technologies combined with a stronger understanding of social behavior are quickly changing these assumptions about security. Automated security systems can prevent a crime before it even happens by recognizing threats before they have a chance to act.
An automated security system combines intelligent cameras with advanced database analysis software. These new intelligent surveillance cameras can automate and streamline security tasks. Smart cameras can track targets as well as make decisions on which targets to watch. These innovative security cameras can key in on individuals and recognize faces and objects the person is carrying. Data can be stored and cross referenced with police databases, hotel registrations, or any other pertinent data to determine the threat level of any given person. As data is collected over time the system is begins to understand how specific people behave. This allows control centers to better monitor a site by requiring fewer agents to monitor more cameras. Utilizing behavioral analysis algorithms the system can notify security monitoring personnel of any suspicious activity.
These systems are astonishingly accurate. One social experiment conducted at MIT, The Reality Mining Project, utilized hundreds of thousands of hours of proximity data collected by tracking mobile phones over a period of nine months. Researchers were then able to create algorithms that could predict a person's next actions accurately over 85% of the time. The program was also able to determine social status and relationships. Tracking individuals as they move through their day from home to work and other key locations the program could create a list of their friends and acquaintances and be right 90% of the time.
How does this work practically speaking? Here is one example. A serious threat, such as a terrorist, is likely to conduct their own surveillance of the site before carrying out an attack. The system could track individuals as they move through the site and recognize surveillance types of behavior. The system would then alert monitoring agents to pay closer attention to them. A snapshot of the target could be used to determine identity and various databases could be cross referenced for any pertinent information about the suspect. They can then direct security teams on how to position themselves to deal with the threat in a moment's notice should the need arise.
